Ironic effects and final target fixation in a penalty shooting task

Human Movement Science 29 (2010) 277–288

Contents lists available at ScienceDirect

Human Movement Science

journal homepage: www.elsevier .com/locate/humov

Ironic effects and final target fixation in a penaltyshooting task

Olaf Binsch a,*, Raôul R.D. Oudejans a, Frank C. Bakker a, Geert J.P. Savelsbergh a,b

a Research Institute MOVE, VU University Amsterdam, The Netherlandsb Institute for Biomedical Research into Human Movement and Health, Manchester Metropolitan University, Manchester, UK

a r t i c l e i n f o a b s t r a c t

Article history:Available online 4 March 2010

PsycINFO classification:23002323233023402346

Keywords:FootballSelf-instructionIronic performanceGaze behaviorVisual search

0167-9457/$ - see front matter � 2010 Elsevier B.doi:10.1016/j.humov.2009.12.002

* Corresponding author. Address: NetherlandsSoesterberg, The Netherlands. Tel.: +31 88 866289

E-mail address: [email protected] (O. Binsch).

The aim of the present study was to find out whether ironic effects ina far aiming task were accompanied by shorter final fixations on thetarget. Generally, it is well known that a sufficiently long final fixa-tion on the target is of crucial importance for accurate performancein far aiming. Recently, it has been shown that ironic effects in golfputts and penalty kicks (in which one does the opposite of whatwas intended, e.g., shoot close to the keeper while attempting toavoid this) were preceded by longer fixations on the to-be-avoidedarea, which may have resulted in shorter final fixations on the targetarea. Therefore, in the current study we examined football playerstaking penalties in a simulated penalty environment with and with-out instructions to avoid the goalkeeper. The findings revealed thatironic effects were indeed accompanied by significantly shorter finalfixations on the target area, i.e., the open goal space. It is concludedthat in far aiming tasks, ironic effects are accompanied by insuffi-ciently long final fixations on the target.

� 2010 Elsevier B.V. All rights reserved.

1. Introduction

In taking penalties in football the persistent wish not to miss may, ironically, increase the likeli-hood that this is precisely what will happen (cf. Bakker, Oudejans, Binsch, & van der Kamp, 2006).In general, the intention to avoid an action may paradoxically increase the tendency to engage in this

V. All rights reserved.

Institute for Applied Scientific Research (TNO), Kampweg 5, 3769 DE,0; fax: +31 34 6353977.

278 O. Binsch et al. / Human Movement Science 29 (2010) 277–288

action (Beilock, Afremow, Rabe, & Carr, 2001; De la Peña, Murray, & Janelle, 2008; Janelle, 1999;Wegner, 1994; Woodman & Davis, 2008). Recent studies in the perceptual-motor domain have shownthat there is a strong relationship between visual attention and performance when ironic effects occur(Bakker et al., 2006; Binsch, Oudejans, Bakker, & Savelsbergh, 2009; Wegner, Ansfield, & Pilloff, 1998).Binsch et al. (2009) showed, for instance, that ironic effects in golf putting were related to changes ingaze behavior. More specifically, they found overall longer average durations of the fixation on the to-be-avoided areas (in front of or behind the hole) and shorter average durations of the fixation on thehole when ironic effects occurred.

Bakker et al. (2006) tested experienced football players shooting penalties with a time constraint of1 s in a simulated penalty environment during differently worded instructions. The task was to shootfootballs towards a screen on which video clips of a goal and goalkeeper were presented for 1 s. Theplayers were asked to shoot as accurately as possible (‘‘accurate condition”), not to shoot within reachof the keeper (‘‘not-keeper condition”), or to shoot to the target area, the open goal space (‘‘open-spacecondition”). Results showed that in the not-keeper condition gaze and shots were more often directedto the to-be-avoided area (the keeper) compared to the accurate and open-space conditions. Bakkeret al. suggested that initial fixations on the keeper may have prevented participants from redirectingtheir attention to the more appropriate area for aiming (open goal space), resulting in more shots atthe keeper.

Still, neither Binsch et al. (2009) nor Bakker et al. (2006) directly examined the role of the finalfixation on the target when ironic effects occurred. It is a well-established fact that the finalfixation on the target prior to and during the final movement (often referred to as ‘‘quiet eye”)is essential in far aiming tasks in general (e.g., Vickers, 1992, 1996; Williams, Singer, & Frehlich,2002; Wilson, Vine, & Wood, 2009; for an overview see Vickers, 2007). In a broad variety of tasks,it has been shown that for accurate aiming a longer final fixation on the target prior to and duringthe aiming action is a characteristic of high levels of skill and accuracy (cf. Vickers, 2007).Furthermore, it appears that emotional factors such as anxiety may lead to shorter final fixationson the target, followed by a decrease in performance (Behan & Wilson, 2008; Vickers & Wil-liams, 2007; Wilson, Vine et al., 2009). The findings by Bakker et al. (2006) that penalty shootersshowing ironic effects spend more time looking at the keeper may imply that, as a consequence,the final fixation on the open-goal space is shorter, which in turn might lead to worseperformance.

In the present study, we examined whether ironic effects in aiming are indeed accompanied bya shorter final fixation on the target. To this aim we investigated football players taking penaltykicks in a similar setting as the one used by Bakker et al. (2006) following negatively and posi-tively worded instructions. More specifically, experienced football players shot penalties, with aconstraint on the response time of 1.5 s, while instructed to shoot as accurately as possible (‘‘accu-rate condition”), to shoot as accurately as possible while being careful not to shoot within reach ofthe keeper (‘‘not-keeper condition”), and to shoot as accurately as possible while being careful toshoot into the open goal space (‘‘open-space condition”). We measured shooting performance, gazebehavior, and response time (time elapsing between presentation of the imperative stimulus, i.e.,the presentation of the projected goal and goalkeeper on a screen and the moment of foot-ballcontact).

Because not all persons tend to show ironic behavior (Beilock et al., 2001; Binsch et al., 2009; De laPeña et al., 2008; Wegner, 1994; Woodman & Davis, 2008), we first distinguished participants who didand did not show ironic effects in the not-keeper condition. Then we compared shooting performance,duration of the initial fixation on the keeper, onset and duration of the final fixation on the open goalspace, and response time across the three instruction conditions (i.e., accurate, not-keeper, and open-space) for those participants who did and who did not show ironic effects. Most importantly, we sub-sequently assessed whether responses indicative of ironic effects were accompanied by longer fixa-tions on the keeper (as was found by Bakker et al., 2006) and shorter final fixations on the opengoal space than responses not showing ironic effects. Given the relevance of final target fixationsfor far aiming and the aforementioned results by Bakker et al. (2006) and Binsch et al. (2009) we ex-pected ironic effects to co-occur with longer keeper fixations and shorter final fixations on the opengoal space.

O. Binsch et al. / Human Movement Science 29 (2010) 277–288 279

2. Method

2.1. Participants

Thirty-two male intermediate football players from a district amateur league (mean age = 21.8 -years, SD = 2.1) participated in the experiment. On average, they had 12.6 years (SD = 4.7) of experi-ence in football competition. At the time of the study all participants were actively engaged infootball competition and they practiced, on average, two times (totaling 3 h) a week. The experimentwas approved by the ethics committee of the Faculty of Human Movement Sciences, VU UniversityAmsterdam. Each participant gave his written informed consent before starting with the experiment.

2.2. Task and design

Video clips of a stationary goalkeeper anticipating a penalty kick were shown on a large screen (seeFig. 1). The clips were made from the perspective of a penalty taker. The task of the participant was totake a penalty without run up towards the projected goal. The player was instructed to hit the screenbefore the projection disappeared, i.e., within 1.5 s after its appearance. Pilot testing had revealed that1.5 s provided players with more than enough time to execute the penalty kicks (Bakker et al. (2006)used a time constraint of only 1.0 s).

The goalkeeper presented in the video clips stood either in the center of the goal, or 0.15 or 0.30 mto the left or to the right from the center, resulting in five clips, one for each position. The off-centerpositions of the keeper were included to induce variation into the shot directions of the players. Eachof the five clips was presented once in each of the three instruction conditions, resulting in five trialsper condition. Within each condition the five trials were randomized. The instructions in each of thethree conditions were as follows: (1) just shoot as accurately as possible: accurate condition; (2) shootas accurately as possible, and be careful not to shoot within reach of the keeper: not-keeper condition;(3) shoot as accurately as possible and be careful to shoot into the open space: open-space condition.

Fig. 1. Experimental set-up: the screen (1) on which the video clips were back-projected with a projector (2), the optical switch(3) that was used to determine when the ball was kicked, the eye-tracker (4) worn by the participant, and the screen camera (5)to determine the shooting performance.

280 O. Binsch et al. / Human Movement Science 29 (2010) 277–288

What it meant to ‘‘shoot as accurately as possible” was explained to the participants prior to testing byshowing each video clip while holding the football (1) close to both the left and the right goal postwhen the goalkeeper stood in the center of the goal, and (2) close to the goal post that was furthestaway from the goalkeeper when the latter stood off-center. In all cases the ball was held close tothe ground.

As we wished to determine a baseline measure (both with respect to performance and gaze behav-ior) unaffected by specific additional instructions, the accurate condition was always presented first.The other two conditions were counterbalanced. Furthermore, following the methodology of Binsch,Oudejans, Bakker, Hoozemans, and Savelsbergh (in press) only trials with the keeper in the center po-sition were analyzed as these provided the most unambiguous determination of (ironic) shootingperformance.

As anticipated, not each participant showed ironic performance in the not-keeper condition: rela-tive to the accurate condition some participants shot closer to the keeper in the not-keeper conditionwhile others did not. Therefore, we distinguished participants who did show ironic effects in the not-keeper condition from those who did not, according to a criterion specified below (see Section 2.5).Therefore, the analytical design of the study also included a group factor (with an ‘‘ironic” and‘‘not-ironic” group).

2.3. Experimental set-up

The video clips (made with a digital video camera, Sony XJ 2000) were back-projected, using a mir-ror and projector, on a large projection screen (2.29 � 2.27 m; see Fig. 1). The projection size of theclips was 2.20 � 1.05 m with a projected goal size of 2.00 � 0.81 m. At each presentation, lasting1.5 s, the player shot a foam ball (Ø = 21, weight: 296 g; thus, with about the same diameter andweight as an official futsal, i.e., indoor football, size 4) to the projected goal from a penalty spot locatedat a distance of 2.83 m from the screen. At that distance the visual angles subtended by the projectedgoal closely resembled the visual angles subtended by the goal for soccer players taking a real life pen-alty from 11 m. A Canon-XM1 video camera, connected to a JVC digital video (DV) recorder and direc-ted at the screen, was used to record shooting performance (at 50 Hz). The video camera wassuspended 2.80 m directly above the penalty spot and aimed at the screen.

Gaze behavior was registered using an eye-tracking system (Applied Science Laboratories 501, Bed-ford, MA) that consisted of a head-mounted scene and (infrared) eye-camera. With the eye-camera,detecting the displacement between the left pupil and cornea reflex relative to a pre-calibrated 9-point grid, the point-of-gaze was determined and integrated into the image of the scene camera thatwas recorded for further analysis. The accuracy of the system was ±1� visual angle. The calibration ofthe system was checked before each trial; if necessary, the system was recalibrated (on average onceper 10 trials). The eye-tracker was connected to the main computer with a 6-m long cable, which wasattached to the waist of the participant but permitted undisturbed shooting mobility.

We used an optical switch located just behind the penalty spot to determine the moment when theball was hit. The optical switch was operated by an infrared light beam that was interrupted by theball when it had been kicked (see Fig. 1). This interruption was converted to a signal that was sentto a light emitting diode (LED) mounted in front of the ASL scene camera (invisible for the participant).Thus, as soon as the ball was hit, the LED turned on and was visible in the scene camera recordings,making it possible to relate gaze behavior to the moment that the ball was kicked.

2.4. Procedure

Participants were tested individually. After a brief general explanation of the experiment partici-pants provided written informed consent. The participant then completed 20 warm-up shots to awhite circle (Ø = 30 cm) projected during 2 s on the black screen at different locations at ground level.After the warm-up, the participant was equipped with the eye-tracker, the system was calibrated andit was explained what it meant to ‘‘shoot as accurately as possible”. Then the participant made 10practice shots to randomly selected clips of goal and keeper, without further instruction. Subsequentlythe trials of the three experimental conditions (three times five shots) were performed. Prior to each

O. Binsch et al. / Human Movement Science 29 (2010) 277–288 281

trial the instruction in question was repeated verbally. Then, a video clip was presented and the par-ticipant kicked the ball towards the goal. After the ball hit the screen, the ball was collected and repo-sitioned on the penalty spot by the experimenter. Hereafter, participants were instructed to look atcertain marks on the screen to verify eye-tracker calibration. Then the next instruction was given,and so on.

2.5. Data reduction

As indicated before, only the trials with the keeper located at the center of the goal were analyzed,thus the resulting data set consisted of 32 (Participants) � 3 (Conditions) � 1 (Shot) = 96 trials. Foreach of these trials we determined shooting performance, that is, the horizontal distance of the ball(in cm) from the center of the goal (keeper) when the ball hit the screen (in short: distance fromthe keeper). In connection with this it should be noted that ball trajectories were almost always closeto the ground; only in six cases the ball landed high in the goal. Single bitmap images were capturedwith Adobe Première 6.5 from the video recordings made by the scene camera. With these images X-coordinates of the landing position of the ball were digitized using Image Digitizing Software ‘‘DIDGE”(cf. Binsch et al., 2009). The distance between the goalposts was used for calibration.

When, in the not-keeper condition, the ball landed at least 10 cm closer to the keeper than in theaccurate condition, this was interpreted as an ironic effect. This criterion, albeit arbitrary, seemed rea-sonable, the difference of 10 cm being half the standard deviation of shooting performance in the accu-rate condition (20.2 cm), and thus substantial. (The results of the analyses with a criterion difference of15 cm proved to be very similar to the results presented below.) To further validate our criterion wealso investigated the number of times that this criterion was met in the open-space condition, in whichby definition ironic effects are out of the question.

Inspection of the video recordings made clear that almost all gaze behavior was within the goalspace (enclosed by the goal posts and bar), although occasionally participants’ gaze was directed atthe floor in front of the goal or outside of the goal. For the purposes of the current study fixations werecoded in terms of three relevant locations within the goal, viz. one keeper area that extended from40 cm to the right to 40 cm to the left of the center of the goal (i.e., the middle 80 cm of the goal)and two open goal spaces, extending 60 cm inwards from each goal post. Together, the latter two areasare referred to as open goal space. We determined the onset and durations of the fixations on the kee-per and the final fixation on the open goal space by analyzing the gaze recordings (frame-by-frame;20 ms per frame) from the first moment during response time that the gaze was directed at the screenuntil the moment at which the football was kicked. In line with the literature the gaze at a specificlocation had to last at least 100 ms, that is, five or more video frames, to be defined as a fixation(cf. Vickers, 1992, 1996, 2007; Williams, Davids, & Williams, 1999).

Finally, in order to check if the expected shorter final fixations on the open goal space in the not-keeper condition might have been due to shorter response times, it was checked whether the responsetimes differed across conditions.

3. Results and discussion

3.1. Number of participants showing ironic performance

There were 14 participants who met the criterion for ironic effect in the not-keeper condition and18 who did not. In the open-space condition there were only two participants who met the criterionused for ironic effect versus 30 who did not. The chi-square test confirmed that the number of timesthat the criterion was met in the not-keeper condition was significantly higher than in the open-spacecondition, v2(1) = 12.0, p < .01. These figures indicate that shots that met the criterion for ironic effectin the not-keeper condition did indeed reflect ironic performance rather than accidental variations inperformance.

282 O. Binsch et al. / Human Movement Science 29 (2010) 277–288

3.2. Shooting performance

First we conducted a 2 (Group: ‘‘ironics” [n = 14], ‘‘not-ironics” [n = 18]) � 3 (Condition: accurate,not-keeper, and open-space) repeated measures ANOVA on shooting performance. This ANOVA re-vealed a significant main effect for condition, F(2, 60) = 6.82, p < .01, g2

p = .19, and no significant maineffect for group, F(1,30) < 2.5, p > .10. Further, as should of course be the case because of the classifi-cation procedure used to distinguish ‘‘ironics” from ‘‘not-ironics”, a significant interaction betweengroup and condition, F(2, 60) = 13.06, p < .001, g2

p = .30, was found (see Fig. 2A and B). Post hoc pair-wise comparisons using Bonferroni correction confirmed that participants who showed ironic perfor-

Fig. 2. Mean shooting distance from the keeper in cm (with SD) for the ‘‘ironic” group (A) and the ‘‘not-ironic” group (B) in thethree instruction conditions. Mean duration of the initial and second fixation on the keeper in ms (with SD; black areas) andmean duration of the final fixation on the open-goal space in ms (with SD; white areas) from the ‘‘ironic” group (C) and the ‘‘not-ironic” group (D) in the three instruction conditions. The times along the y-axis are presented relative to the zero-axis indicatingthe moment that the ball was kicked. (Note that the mean durations of the second fixation on the keeper may be less than100 ms because many participants did not fixate the keeper for a second time; see text for explanation.)

O. Binsch et al. / Human Movement Science 29 (2010) 277–288 283

mance in the not-keeper condition shot 24.3 cm closer to the keeper in this condition compared to theaccurate condition, p < .01 (Fig. 2A). Moreover, these participants also shot 24.6 cm closer to the keeperin the not-keeper condition than in the open-space condition, p < .01, a difference that played no role inthe classification procedure. By comparison, there were no significant differences in shooting distancefrom the keeper among the three conditions for participants who showed no ironic performance,ps > .17 (Fig. 2B).

3.3. Gaze behavior

Figs. 2C and D show average durations of fixations on the keeper and on the open goal space untilthe ball was kicked in each of the conditions for those participants who showed ironic effects (Fig. 2C)and for those who did not (Fig. 2D). As can be seen, there were two general patterns of gaze behavior.In all cases participants first fixated the keeper followed by a final fixation on the open goal space. Inmost cases this final fixation on the open goal space lasted until the ball was kicked (e.g., the accurateand open-space conditions in Fig. 2C). In a number of cases participants returned their gaze to the kee-per before the ball was kicked (e.g., not-keeper condition in Fig. 2C). To test whether ironic effects inthe not-keeper condition (see Fig. 2A) were accompanied by either a longer first fixation duration onthe keeper, an earlier onset or a shorter duration of the final fixation on the open goal space, or a com-bination of these, 2 � 3 mixed design ANOVAs were conducted with the factors group (‘‘ironics” and‘‘not-ironics”) and condition (accurate, not-keeper, and open-space) and with repeated measures on thelast factor.

3.3.1. Duration of the initial fixation on the keeperThe first 2 � 3 ANOVA on duration of the fixation on the keeper before the final fixation on the open

goal space revealed neither a significant main effect for group or condition, nor a significant interac-tion between group and condition, Fs < 2.5, ps > .10, indicating that the ironic performance was notaccompanied by a longer initial fixation duration on the keeper (see Fig. 2C and D).

3.3.2. Onset of the final fixation on the open goal spaceThe Group � Condition ANOVA on the onset of the final fixation on the open goal space revealed a

trend for group, F(1, 30) = 3.52, p = .071, g2p = .10, and a main effect for condition, F(2, 60) = 3.62,

p < .05, g2p = .11, in the absence of a significant interaction between group and condition, F(2,

60) < 1.5, p > .25. Post hoc pair-wise comparisons with Bonferroni correction revealed that the onsetof the final fixation on the open goal tended to be earlier in each condition for the participants whoshowed ironic effects (M = 214 ms before ball contact) compared to participants who did not show iro-nic effects (M = 225 ms before ball contact), p = .071. Furthermore, participants also showed an earlieronset of the final fixation on the open goal space in the accurate condition (M = 273 ms before ball con-tact) compared to the open-space condition (M = 208 ms before ball contact), p < .05.

3.3.3. Duration of the final fixation on the open-goal spaceThe ANOVA on the duration of the final fixation on the open goal space revealed a trend for group,

F(2, 60) = 3.79, p = .061, g2p = .11, a significant main effect for condition, F(2, 60) = 6.18, p < .01, g2

p = .17,as well as a significant interaction between group and condition, F(2, 60) = 6.41, p < .01, g2

p = .18 (seeFigs. 2C and D). Post hoc pair-wise comparisons using Bonferroni correction showed that for partici-pants who had shown an ironic effect in the not-keeper condition the average final fixation on the opengoal space (M = 129 ms, SD = 68) in the not-keeper condition was significantly shorter than in the accu-rate (M = 224 ms, SD = 79) and the open-space (M = 206, SD = 56 ms) conditions, respectively (ps < .01,see Fig. 2C). Thus, ironic performance was accompanied by shorter final fixations on the open goalspace. Participants who did not show ironic performance did not have significantly shorter final fixa-tions on the open goal space in the not-keeper condition compared to the other two conditions,ps > .13. The latter participants fixated the open goal space 39 ms longer in the accurate conditioncompared to the open-space condition, p < .05. Moreover, participants who did not show ironic shotsfixated the open goal space 99 ms longer in the not-keeper condition than participants who did showironic effects in this condition, p < .01 (see Fig. 2C and D).

284 O. Binsch et al. / Human Movement Science 29 (2010) 277–288

3.4. Response time

To check whether the shorter fixations on the open goal space that accompanied ironic effects wererelated to shorter response times, a Group � Condition ANOVA on response time was performed.There was a significant main effect for condition, F(1, 30) = 15.99, p < .001, g2

p = .35, but no significantmain effect for group, F(1, 30) < 0.50, p = .49, and no significant interaction between condition andgroup, F(2, 60) < 0.50, p = .66. Post hoc pair-wise comparisons with Bonferroni correction revealed thatthe main effect for condition was due to the fact that the participants used more time for task execu-tion in the open-space condition (M = 1037 ms, SD = 206) compared to the accurate (M = 845 ms,SD = 158) and not-keeper conditions (M = 893 ms, SD = 208), ps < .01. Importantly, the response timesof the participants who showed ironic performance were not significantly shorter than those of theparticipants who did not show ironic performance. Thus, the shorter final fixation on the open goalspace that accompanied ironic performance was not the result of reduced response times.

In sum, it seems probable that the final fixation on the open goal space in the not-keeper conditionwas not sufficiently long for participants who showed ironic performance in this condition. As a resultof this short final fixation they seemed to have shot closer to the keeper. In principle, these shorterfinal fixations on the open goal space might have been due to a subsequent return of the gaze tothe keeper before the ball was kicked (see Fig. 2C, not-keeper condition), as well as to a long initial fix-ation on the keeper (even though on average the latter fixations in the not-keeper condition did notsignificantly differ between the participants who showed ironic performance and those who did not).

To shed light on this issue we split up the ironic group into those participants who showed an addi-tional fixation on the keeper after the final fixation on the open goal space just before the ball waskicked (n = 8) and those who did not (n = 6). Fig. 3 shows average shooting performance (top panels)as well as accompanying gaze behavior (bottom panels) for these two ‘‘ironic” subgroups. Duration ofthe initial fixation on the keeper, and the onset and the duration of the final fixation on the open goalspace were further analyzed for these two subgroups. Shooting performance and response time werealso further analyzed.

However, these latter two analyses yielded no significant differences between the subgroups, indi-cating that both subgroups showed ironic effects to the same degree and that these effects were notrelated to response time. Therefore, these analyses are not reported in detail (see Fig. 3 top panels forshooting performance).

3.5. Comparison of the two ‘‘ironic” subgroups

3.5.1. Duration of the initial fixation on the keeperA 2 (Subgroup: ‘‘ironics” with extra fixation on the keeper [n = 8], ‘‘ironics” without extra fixation

on the keeper [n = 6]) � 3 (Condition: accurate, not-keeper, and open-space) mixed design ANOVA onthe average duration of the initial fixation on the keeper revealed no significant effect for group, nosignificant effect for condition nor a significant interaction between group and condition, Fs < 0.75,ps > .50, indicating that initial fixations on the keeper were of comparable duration for both subgroupsin all three conditions (see Figs. 3C and D).

3.5.2. Onset of the final fixation on the open-goal spaceThe Group � Condition ANOVA conducted on the onset of the final fixation on the open goal space

revealed a trend for group, F(1, 12) = 3.64, p = .081, g2p = .23, but no significant effect for condition, F(1,

12) < 1.50, p = .33. Furthermore, there appeared to be a significant interaction between group and con-dition, F(2, 24) = 5.26, p < .05, g2

p = .31. Post hoc pair-wise comparisons with Bonferroni correction re-vealed that the onset of the final fixation on the open goal space was significantly later in the not-keeper condition for participants who had no extra fixation on the keeper in this condition(M = 87 ms before foot-ball contact, see Fig. 3D) compared to the participants who did have an extrafixation on the keeper (M = 278 ms before foot-ball contact, see Fig. 3C), p < .01.

Fig. 3. Mean shooting distance from the keeper in cm (with SD) for the ‘‘ironic” subgroup with (A) and without (B) a secondfixation on the keeper in the not-keeper condition in the three instruction conditions. Mean duration of the initial and secondfixation on the keeper in ms (with SD; black areas) and mean duration of the final fixation on the open-goal space in ms (withSD; white areas) for the ‘‘ironic” subgroup with (C) and without (D) a second fixation on the keeper. (Note that the meandurations of the second fixation on the keeper may be less than 100 ms because many participants did not fixate the keeper fora second time; of course, the mean duration of this second fixation on the keeper equals 0 ms for the ‘‘ironic” subgroup in thenot-keeper condition that did not fixate the keeper for a second time in this condition).

O. Binsch et al. / Human Movement Science 29 (2010) 277–288 285

3.5.3. Duration of the final fixation on the open-goal spaceThe Group � Condition ANOVA conducted on the duration of the final fixation on the open goal

space revealed a significant main effect for condition, F(2, 24) = 7.67, p < .01, g2p = .39, in the absence

of a main effect for group and in the absence of a significant interaction between group and condition,Fs < 0.75, ps > .50. Post hoc pair-wise comparisons only confirmed that for participants who showedironic performance the final fixation on the open goal was shorter in the not-keeper condition com-pared to the accurate and open-space conditions, ps < .01. Importantly, there was no significant differ-ence in the duration of the final fixation on the open goal space between those who did and did nothave an extra fixation on the keeper before the ball was kicked (see Figs. 3C and D).

286 O. Binsch et al. / Human Movement Science 29 (2010) 277–288

4. General discussion

The aim of the present study was to find out whether ironic effects in far aiming tasks are accom-panied by shorter final fixations on the target, as would be expected on the basis of the literature (Bak-ker et al., 2006; Binsch et al., 2009). To this aim we tested experienced football players shootingpenalties in an experimental penalty indoor setting under differently worded instructions. Fourteenof the 32 participants (44%) showed ironic performance in the not-keeper condition, which is morethan reported in several other studies (Binsch et al., 2009; De la Peña et al. 2008; Woodman & Davis,2008). Most importantly, the combined results confirm that ironic performance was closely related toan insufficiently long final fixation on the open goal space (the target) whether or not this fixation wasfollowed by an additional fixation on the keeper. For the ‘‘ironic” group (and subsequently both‘‘ironic” subgroups) the duration of the final fixation in the not-keeper condition was significantlyshorter than in the other conditions. When performance was not-ironic the minimal average durationof the final fixation on the open goal space was always about 200 ms. In short, ironic performance wasaccompanied by a (too) short final fixation on the open goal space, the most appropriate gaze locationfor successful aiming in this experimental setting, either because ‘‘ironic” participants disengagedtheir gaze from the keeper later, that is, closer to kicking the ball, or because they showed an extrafixation on the keeper prior to kicking the ball. Earlier findings that ironic effects occur when the gazedwells longer on the to-be-avoided area (e.g., the goal keeper; Bakker et al., 2006; Binsch et al., 2009),already suggested that as a consequence fixations on the target area might be shorter. The presentstudy is the first one in which this relationship between the duration of the final fixation on the targetand the occurrence of ironic effects has been explicitly shown.

An explanation for the shorter final fixation on the target might be that, following a negativeinstruction, participants hasten their performance leaving insufficient time to fixate the target areaafter attention is first drawn to the keeper. However, as was shown here, response times were notshorter in the not-keeper condition (when ironic effects occurred) in comparison to the accurate con-dition. Of course, it should be noted that in the present study the participants had ample time to exe-cute their shots; the response times varied from 845 through 1037 ms (ball flight not included),whereas 1.5 s were available. In practice, time constraints may play a role in ironic effects, especiallyas performers tend to hasten their performance under pressure (cf. Jordet, 2009; Nieuwenhuys &Oudejans, 2009).

Two phenomena responsible for the shorter final fixations on the target area (open goal space) re-lated to the ironic effect were demonstrated in the present study. The first one, occurring in 6 of the 14participants showing ironic effects, implies that the initial fixation on the keeper lasted too long as toleave enough time for subsequent fixation of the open goal space. The second one, occurring in theremaining eight participants showing ironic effects, implies that the final fixation of the open goalspace did not last long enough because the gaze returned once more to the keeper. Both phenomenacan be explained by Wegner’s (1994) theory of ironic processes, which holds that successful behaviorrelies on two cognitive processes: one controlled and the other automatic. In brief (for more elaboratedescriptions we refer to Wegner; cf. Janelle, 1999), the automatic search process continuously scansthe contents of consciousness for any trace of unwanted thoughts. When an unwanted thought is de-tected, the controlled system ‘‘kicks in” and replaces this item with a more appropriate task-relatedthought. Under certain circumstances, for instance, in situations with a high mental load or with timepressure, the controlled replacing process, which requires attention for successful initiation, can becompromised – resulting in the manifestation of unwanted thoughts and less than-optimal perfor-mance. In the present study, the negative instruction not to aim within reach of the keeper may havecaused the ‘‘keeper” to linger on in the cognitive system. In one case this may have led to difficulty indisengaging from the keeper, leading to a too short final fixation on the target. In the second case, inwhich the ‘‘ironic” participants showed an extra fixation on the keeper, they apparently remainedmore easily distracted by the keeper even after their gaze had already moved to the open goal space.So even though attention was already directed to the open goal space, the keeper apparently lingeredon in the cognitive system, once more drawing (visual) attention and thus leaving insufficient time fora proper final fixation on the target.

O. Binsch et al. / Human Movement Science 29 (2010) 277–288 287

Interestingly, both the inability to disengage from the keeper, who is a potential distracter in thepenalty kick task, and the enhanced distractibility (by the keeper), closely resemble the effects of anx-iety on perceptual-motor performance. Comparable results, that is, shorter final fixations on the targetand/or more or longer fixations on potential distracters (such as the keeper in penalty shooting), werefound in archery, handgun shooting, rifle shooting in biathlon, free throw shooting and penalty shoot-ing under high anxiety conditions (Behan & Wilson, 2008; Binsch et al., in press; Nieuwenhuys &Oudejans, 2009; Vickers & Williams, 2007; Wilson, Vine et al., 2009; Wilson, Wood, & Vine, 2009).Comparison of the effects of anxiety in the latter studies with the ironic effects found in the presentstudy, as well as in the studies by Bakker et al. (2006) and Binsch et al. (in press), makes clear thatanxiety and ironic instructions can have similar effects on gaze and aiming behavior. Recently, the ef-fects of anxiety on cognitive as well as on perceptual-motor performance have been explained withinthe framework of attentional control theory (Eysenck, Derakshan, Santos, & Calvo, 2007; Nieuwenhuys& Oudejans, 2009). In brief, this theory holds that anxiety leads to changes in attention and attentionalcontrol involving a shift from the more controlled goal-directed attentional system (top down), influ-enced by action goals and intentions, to the more automatic stimulus-driven attentional system (bot-tom up) geared to all kinds of stimuli (e.g., threat-related stimuli) that might affect reaching thosegoals (Corbetta & Shulman, 2002; Eysenck et al., 2007). More in particular, these changes imply botha higher distractibility by task-irrelevant (threat-related) stimuli like the keeper in the present settingfollowing a negative instruction, and a greater difficulty to disengage from such stimuli. In short, thereare striking resemblances between ironic effects and effects of anxiety on perceptual-motor perfor-mance, as well as between the self-regulatory mechanisms proposed in Wegner’s (1994) theory of iro-nic mental processes and in attentional control theory (Eysenck et al., 2007; cf. Wilson, Vine et al.,2009). Both theories are dual-process theories in which automatic and controlled processes are sup-posed to interact, and in which the interaction may be affected by circumstances leading to subopti-mal performance. It is a challenge for future research to gain more insight into the differences andsimilarities between the effects of ironic instructions and anxiety on performance in terms of boththeories.

With regard to practice our findings should not be taken to imply that in actual penalty kicking thesame results would have been found. On a football pitch several important constraints are differentfrom those employed here (time constraints, pressure; e.g., Jordet, 2009; Jordet, Elfering-Gemser,Lemmink, & Visser, 2006). Thus, the generalizability of our findings to on-field penalty taking shouldbe viewed with caution. Still, in far aiming tasks it seems important to avoid negative instructionsinvolving the to-be-avoided area (e.g., the goal keeper), either by coaches or instructors or by athletesthemselves. The present results also show that positive instructions involving the target area (open-space instruction) led to sufficiently long final fixations on the target, a somewhat extended task dura-tion and good performance. Therefore, such positive instructions should be favored (cf. Binsch et al., inpress). In agreement with earlier findings, the results of the present study once more emphasize theimportant role of the final fixation on the target in far aiming, this time in relation to ironic effects.

References

Bakker, F. C., Oudejans, R. R. D., Binsch, O., & van der Kamp, J. (2006). Penalty shooting and gaze behavior: Unwanted effects ofthe wish not to miss. International Journal of Sport Psychology, 37, 265–280.

Behan, M., & Wilson, M. (2008). State anxiety and visual attention: The role of the quiet eye period in aiming to a far target.Journal of Sports Sciences, 25, 207–215.

Beilock, S., Afremow, J. A., Rabe, A. L., & Carr, T. H. (2001). ‘‘Don’t miss!” the debilitating effects of suppressive imagery on golfputting performance. Journal of Sport and Exercise Psychology, 23, 200–221.

Binsch, O., Oudejans, R. R. D., Bakker, F. C., Hoozemans, M. J. M., & Savelsbergh, G. J. P. (in press). Ironic effects in a simulatedpenalty shooting task: Is the negative wording in the instruction essential? International Journal of Sport Psychology.

Binsch, O., Oudejans, R. R. D., Bakker, F. C., & Savelsbergh, G. J. P. (2009). Unwanted effects in aiming actions: The relationbetween gaze behavior and performance in a golf putting task. Psychology of Sport and Exercise, 10, 628–635.

Corbetta, M., & Shulman, G. L. (2002). Control of goal-directed ad stimulus-driven attention in the brain. Nature ReviewsNeuroscience, 3, 201–215.

De la Peña, D., Murray, N. P., & Janelle, C. M. (2008). Implicit overcompensation: The influence of negative self-instructions onperformance of a self-paced motor task. Journal of Sports Sciences, 26, 1323–1331.

Eysenck, M. W., Derakshan, N., Santos, R., & Calvo, M. G. (2007). Anxiety and cognitive performance: Attentional control theory.Emotion, 7, 336–353.

Janelle, C. M. (1999). Ironic mental processes in sport. The Sport Psychologist, 13, 201–220.

288 O. Binsch et al. / Human Movement Science 29 (2010) 277–288

Jordet, G. (2009). Why do English players fail in soccer penalty shootouts? A study of team status, self-regulation, and chokingunder pressure. Journal of Sports Sciences, 27, 97–106.

Jordet, G., Elfering-Gemser, M. T., Lemmink, K. A. P. M., & Visser, C. (2006). The ‘‘russian roulette” of soccer? Perceived controland anxiety in a major tournament penalty shootout. International Journal of Sports Psychology, 37, 281–298.

Nieuwenhuys, A., & Oudejans, R. R. D. (2009). Effects of anxiety on handgun shooting behavior of police officers: A pilot study.Anxiety, Stress, and Coping. First published online 22 May 2009 (iFirst), doi:10.1080/10615800902977494.

Vickers, J. N. (1992). Gaze control in putting. Perception, 21, 117–132.Vickers, J. N. (1996). Visual control when aiming at a far target. Journal of Experimental Psychology: Human Perception and

Performance, 22, 342–354.Vickers, J. N. (2007). Perception, cognition, and decision training: The quiet eye in action. Champaign, IL: Human Kinetics.Vickers, J. N., & Williams, A. M. (2007). Performing under pressure: The effects of physiological arousal, cognitive anxiety, and

gaze control in biathlon. Journal of Motor Behavior, 39, 381–394.Wegner, D. M. (1994). Ironic processes of mental control. Psychological Review, 101, 34–52.Wegner, D. M., Ansfield, M., & Pilloff, D. (1998). The putt and the pendulum: Ironic effects of the mental control of action.

Psychological Science, 9, 196–199.Williams, A. M., Davids, K., & Williams, J. G. (1999). Visual perception and action in sport. London: Taylor and Francis.Williams, A. M., Singer, R. N., & Frehlich, S. G. (2002). Quiet eye duration, expertise, and task complexity in near and far aiming

tasks. Journal of Motor Behavior, 34, 197–207.Wilson, M. R., Vine, S. J., & Wood, G. (2009). The influence of anxiety on visual attentional control in basketball free throw

shooting. Journal of Sport and Exercise Psychology, 31, 152–168.Wilson, M. R., Wood, G., & Vine, S. J. (2009). Anxiety, attentional control, and performance impairment in penalty kicks. Journal

of Sport and Exercise Psychology, 31, 761–775.Woodman, T., & Davis, P. (2008). The role of repression in the incidence of ironic errors. The Sport Psychologist, 22, 183–196.